Very interesting proposal, although I’d like to point out a very essential element: your references are combat games, where it is generally agreed that the point is to kill enemies. In Thrive, the fight is not always a given! One may, as an example, decide to flee heavies; but at the same time, speedies might flee the player instead. Conversely, cells may attack unexpectedly.
This has to be kept in mind: enemies not only have to be defeated, but they also have to be engaged (or not). If no interesting engagement happens, it will be boring to the player. Conversely, if the player can’t help and is always engaged, it might be a difficult situation (but that would be e.g. the main activity in plant gameplay).
Small remarks on the rest:
- membranes can offer specific damage reduction, so that can count as shielded;
- a viable evo strategy could be active scavenging, that is helping another cell (e.g. through mucilage) to get part of their food; pack mechanisms could be of interest here (even though it isn’t properly a pack as there are different species);
- don’t cells with toxins shoot some upon death? maybe it was patched;
- bombers exist, e.g. bees are evolved to die when fighting; but that is not really something I can imagin in bacteria as they have little “kinship awareness”;
- I foresee this resonating with @NickTheNick’s latest proposal for auto-evo